萊迪思半導體 (LSCC) 2005 Q1 法說會逐字稿

Good morning and welcome to today's conference call.

Copies of the Lattice Semiconductor first-quarter ending March 31, 2005 earnings press release may be obtained from the Company's Website, which is www.lscc.com.

This call is being recorded and it is being broadcast live over the Internet by CCBN.

A live broadcast and replay of the call will be available on the Lattice Investor Relations Website, www.lscc.com.

At this time, I would like to turn the call over to Chief Financial Officer, Jan Johannessen.

Please go ahead, sir.

Thank you.

And good morning, everyone.

Joining me on the call today are Cyrus Tsui, our CEO, Steve Skaggs, our President, and Rodney Sloss, our Vice President of Finance.

Before we begin, I would like to read a Safe Harbor statement and then give a financial review of the first quarter.

Then Steve will provide a business review followed by our second-quarter outlook.

We will then hold a question and answer section.

I will now read the Safe Harbor statement.

This conference call may contain forward-looking statements within the meaning of the Federal securities laws including statements about our future quarterly financial results, revenues, customers, product offerings, and our ability to compete.

Investors are cautioned that actual events and results could differ materially from these statements as a result of a number of factors including; general economic conditions, overall semiconductor market conditions, market acceptance and demand for our new products, our dependencies on our silicon wafer suppliers, the impact of competitive products and pricing, technological and product development risks.

We do not intend to update or revise any forward-looking statements.

Now to the financial review.

Revenue for the first quarter was $51.3 million, up 6% sequentially from the fourth quarter and above our original forecast of $46 million to $49 million.

Proceeding with the rest of the statement of operations.

Gross margin for the quarter was 56.8% up slightly from the 56.6% posted in the fourth quarter.

Quarterly R&D expense was $24.6 million, up $1.7 million from the prior quarter and in line with our guidance.

Quarterly SG&A expense was $14.3 million, up $1.3 million from last quarter.

The increase was primarily a result of the incremental expense accrual relating to legal fees for our securities litigation.

Other income for the March quarter was $3.4 million up $2.3 million from the fourth quarter.

The increase was primarily the result of buying back $16.5 million of our convertible bonds during the first quarter.

We added $100,000 to our tax provision for foreign taxes during the first quarter.

Intangible asset amortization was $4.4 million for the quarter, down $1.3 million down from the fourth quarter.

Intangible asset amortization for the second quarter will be about $4.1 million.

And the total for 2005 will be about $16.2 million and for 2006 about $10.8 million.

This compared to a total of $124 million of amortization expenses for 2000 and 2004. the amortization of intangible assets will be substantially eliminated in 2008.

The March quarter net loss was $10.9 million or $0.10 per share, an improvement from net loss of $13.1 million or $0.12 per share for the prior quarter.

These losses include charges of $4.4 million and $5.8 million respectively for the amortization of intangible assets.

On a non-GAAP basis which excludes the aforementioned intangible asset amortization, we had a net loss of $6.5 million or $0.06 per share, an improvement of a net loss of $4.7 million or $0.07 per share posted in the fourth quarter.

Now turning to the balance sheet.

Cash and short-term investments at March 2005 were $253 million.

During the first quarter, we made the second $25 million installment of the advanced payment on wafer purchases to Fujitsu under the technology development agreement announced last year.

This is the second of four payments totaling $125 million that will be paid based on achievement of specific milestones.

We currently expect to make the two final payments totaling $75 million in the first half of 2006.

Accounts receivable increased to a more normal level of 24.9 million on March 31, up from 19.6 million in the prior quarter.

And days sales outstanding increased to more a normal 44 days from 37 days.

We spent $1.9 million on capital expenditures during the first quarter and depreciation expense was $3.8 million.

Inventory declined $2.1 million to $36.5 million and now stands at about 4.9 months on a cost of sales basis, which is within our target range of 4 to 5 months.

We have thus been able to successfully work down the excess inventory to a normal level without taking any extraordinary writedowns.

The high level of inventory on hand was the result of the high-tech bubble busting in 2001.

As I mentioned earlier, we used cash to repurchase $16.5 million to our convertible bonds during the quarter.

Collections for the first quarter were lower than normal after markets and channels worked off their excess inventories in the quarter.

Coupled with an extra payroll in the first quarter, we experienced a negative cash flow of 2.9 million for first quarter excluding $25 million Fujitsu payments.

We expect a positive operating cash flow in the second quarter as accounts receivable balances return to normal levels.

To conclude, I would like to point that we have over the past three years generated more than $100 million from cash from operations before investments and have repurchased over $100 of convertible debt.

We have today over $350 million in cash liquid securities from prepaid wafers.

Inventories are back to target levels.

The majority of the intangible asset amortization is behind us.

And our convertible debt level has been reduced by over 40%.

As a result, we are today financially stronger than we were three years ago despite the heavy investments we have been making in R&D.

This concludes the financial review portion of the call.

I would now like to turn the call over to Steve Skaggs.

Thanks, John.

During the first quarter, we saw a return to sequential quarterly revenue growth as we experienced improved market conditions, an acceleration of growth from our new products, and a resumption of growth from our mainstream products.

Geographically during the quarter the Americas made up 30% of revenue.

Europe, 25%.

And Asia, 45%.

Growth last quarter was led by Asia.

Revenue by end market for the quarter was as follows: Communications made up 53% of our revenue; computing, 19%; and industrial/other, 28%.

During the first quarter, FPGA product revenue was 9.2 million or 18% of revenue and grew 4% sequentially.

This growth was led by our FPSC products which posted record quarterly revenues and compensated for declines in our FPGA products.

PLD product revenue accounted for 42.0 million or 82% of our review and grew 6% sequentially.

This growth was driven by our new CPLD products.

We posted our ninth consecutive quarter of sequential revenue growth from our new product categories as historic customer designs continue to move into volume production during the first quarter.

New products grew 13% sequentially and now make up 28% of total revenue.

When looked at on a year-over-year basis, new product revenue grew 86%.

Just to remind you of these new products basically consist of devices that went into production during the 2000 to first half of 2004 time frame and include broadly in the PLD segments our ispMACH 4000 CLPD's, as well as our mixed signal products.

And in the FPGA segment, our FPSC and ispXPGA families.

The revenue growth we see today reflects customer design that was generated in 2003/2004 time frame that only now are moving into volume production.

During the fourth quarter, overall - - during the first quarter, excuse me - - overall customer design ends of our new products established a new record and grew 12% sequentially.

We expect continued revenue growth from new products as our historic customer designs continue to move into volume production.

Revenue from mainstream products grew 10% sequentially and now account for 36% of total revenue.

As you'll recall, these products experienced a significant sequential revenue decline during the fourth quarter of calendar 2004, which we attributed to an inventory correction.

The reversal of that negative revenue trajectory during first calendar quarter and a return to positive sequential growth is a positive sign as it indicates this excess customer and channel inventory has been worked off.

Finally, mature products declined 3% sequentially and now make up 35% of our total quarterly revenue.

Operationally, the management team and a majority of our resources remain strongly focused on new product development efforts which are intended to prove our position in the large and attractive FPGA market.

Industry guard, I would like to now update you on what we have accomplished in this area since our last quarterly conference call.

Last quarter, we introduced our second planned next generation FPGA platform, the LatticeXP non-volatile FPGA family.

Manufactured using 130 nanometer embedded flash process jointly developed with our new foundry partner, Fujitsu.

This family provides Lattice a strongly differentiated entry into the FPGA market.

The LatticeXP family is targeted to the sweet spot of the low cost FPGA segments and offers larger capacities of between 3 and 20,000 look-up tables.

Additionally this innovative family offers customers compelling advantages in the area of design security, instant on-logic functionality, and field reprogrammability when compared to the SRAM bases, volatile FPGA's that currently make up over 90% of the market.

At present, we are sampling the first member of this family and have already secured our first customer designs.

We also completed the volume production release of our first next generation FPGA platform, the Lattice EC and LatticeECP product families.

Just to refresh everyone on the timeline, these product families were first publicly announced at the beginning of the third quarter of 2004.

We shipped our first device samples during that same quarter, and by the end of last calendar year had sampled 10 of the 12 unique silicon devices in these families.

At present, all 12 devices are fully qualified.

Are released to volume production and have shipped to our customers.

Now to go from first device samples to full-production release of a major new product family in less than nine months is really quite an accomplishment.

And reflects the strong execution of our product development team working in conjunction with our foundry partner, Fujitsu.

On the software front, early in this current week, we released version 5.0 of the ispLEVER development tool.

Our third major software release in the last 11 months.

This latest version provides customer access to both the Lattice EC, LatticeECP and the Lattice XPGA families.

In addition to delivering significant functional and performance upgrades.

Based on our own internal benchmarks, device utilization improves by 6% while run time has been cut by 14%.

All figures are compared to the last release of our software.

We are also pleased to announce the renewal of our OEM agreement with Synplicity and the new developments and marketing agreements that will bring us enhanced support under Synplicity's most advanced FPGA synthesis tool, enhanced tool, SynplifyPro.

We have made strong progress on the software front since our initial entry into the FPGA market.

And at this stage, I believe we have substantially closed the competitive gap.

Last quarter in this forum I mentioned we would provide you with a quantification of our customer design activity in order to allow you to better gauge the market reception of these important new FPGA products.

As of the end of the first quarter, we have secured over 200 design end at approximately 150 distinct customers for our new FPGA products.

Additionally at this early stage we have booked over $100,000 in revenue to support these designs.

I should also note that for these new FPGA products, we have established what we believe is a fairly rigorous definition of a customer design end.

To be counted in a metric I just reported, a customer must have implemented the design end in our ordered and received silicon, and placed the Lattice FPGA device on a prototype board.

Although we remain encouraged about the market potential for our new generation FPGA's, I still must caution you that at this stage it is still premature to make a revenue forecast.

We will however, continue to keep you updated in this forum on our new FPGA design activities.

I would like now to turn to a discussion of our second-quarter outlook.

As I mentioned at the beginning of my remarks, the first quarter saw a general rebound in the PLD market.

During the second quarter we expect market conditions will remain favorable and we anticipate continued revenue growth led by our new products.

Our current estimate is for sequential quarterly revenue growth in the low to mid single digits on a percentage basis.

For the rest of the P&L, we currently have the following expectations for the June 2005 quarter.

We expect gross margin as a percentage of revenue to be essentially flat.

Operating expenses will be - - continue to be dependent on new product development activities; however are expected to decrease by approximately $1 to $2 million during the second quarter.

We expect intangible asset amortization, as Jan mentioned, to be approximately $4.1 million.

We expect approximately $1.5 million of other income.

We'll continue to report approximately $100,000 of tax expense.

And finally, we expect the share count to be relatively flat.

And with that I'd like now to open the call for questions.

Operator we can begin the Q&A portion of the call.

Thank you. [Operator Instructions] Let's go to Chris Danely from J.P.

Morgan with our first question.

Thanks, guys.

Excuse me.

Can you just talk about - - sorry I am a little sick - - can you just talk about the book-to-bill and what the turns percentage required to hit the guidance?

And how has April been so far?

Sure.

Book-to-bill last quarter was over 1.

We don't disclose the exact figure.

Turns last quarter were approximately 70%, which was higher than our initial expectation and accounts for the higher end guidance revenue.

The turns requirement for the guidance I gave is in the mid-60s range for the second quarter.

Great.

And also can you just comment on what you've seen out there in the end markets, whether you expect relative strength or you see anything that has gotten a little more challenging?

Sure.

You know, if you look last quarter on a percentage of revenue basis, the communications end market grew really 5 percentage points for us.

We saw some strength in the wireless market and also in the ASIC segment.

But generally the market was healthy across the board as really inventory drawdowns ended.

And the communication market being the largest percentage of our business really reflects the overall trend that we saw there.

Computing for us was relatively stable.

Last quarter was strength in the server and storage areas essentially offset by some weakness in the peripheral areas.

And the other end markets showed some decline really that was attributable to weakness in the industrial, mainly the test sector and the military segment that reflects some product transitions - - at the customers' specific product transitions.

And there was some seasonal weakness in the consumer markets as the first quarter tends to be a seasonally soft period for that business.

Going forward, we would expect general health in the communication markets to continue.

And the computing market to be relatively stable, and I would expect the industrial other market to continue its long-term trend, which is to increase consumption of PLD products.

Great and then last question.

Will you guys continue to buy back converts going forward?

You know, we do this opportunistically if we get some offers at attractive prices.

So it depends what the prices are.

Okay.

Thanks.

Our question comes from Mark Edelstone with Morgan Stanley.

Good morning, guys.

A couple of questions.

First off, can you break off what the gain was in the quarter for the convertible purchases?

Yes, I can do that for you.

The - - just hold on a second.

Do you have a second question and I can take that while you are on Jan.

Yes, I guess either for you or Cyrus.

With the snap-back that you saw here in Q1, do you have a sense of how much of that was driven by the end of the inventory corrections?

So in other words, at this stage, do you have a sense as to where your shipments were in Q1 relative to true end consumption of your products?

We believe that consumption - - shipments in consumption are in line, Mark.

And the new product growth is really driven by new volume production and is less sensitive to inventory corrections.

The snapback in the mainstream products reflects the inventory rebalancing that occurred.

You know we haven't done a detailed calculation to break out our growth by those two factors.

Going forward, we would expect, in more normal industry conditions to prevail.

And our new products to continue to grow.

And that's reflected in the guidance I gave.

So it's fair to suggest then that if you look at the mainstream products for Q2 and beyond, the performance there is going to be directly related to end market demand as opposed to any further snap back after the inventory correction?

Yes.

We've said all along that the mainstream products are the products that are most sensitive to inventory trends in the industry.

You know mature products are more likely to secularly decline.

New products, if we are doing our job, should be growing and relate to design opportunities.

Mainstream products are the most susceptible to industry rebalancing industry and tend to move reflective of those trends.

Hey, Mark.

While Jan is looking that up, just one other question as well.

You gave us the details of how you are defining design wins for the new products.

Is that the exact same definition you have used in the past?

It is more rigorous and it's something that we were able to validate shipment of silicon on a per customer basis because the numbers are much smaller than our CPLD business.

So it's a much more rigorous definition.

With the CPLD business we have tended to rely on the sales force' basic statement that the customer has specified our product and secured a socket.

We don't pay people on design wins.

We pay people on design ends, which is shipment to design socket.

The definition of design win for FPGA has been made much more rigorous because we want to report it to the financial community, and we want to be able to in the future to be able to formulate a revenue outlook based upon those design activities.

Okay.

But I guess using your old definition, it's just not apples and apples.

You can't use the history of how long it would have taken from that design end to get to volume production or what kind of lifetime revenues came off of that design win.

It just wouldn't be comparable with what you are doing here for FPGA's, it sounds like.

Yes, that's correct.

And the CPLD business for us has a different dynamic since we are already typically engaged with customers with designs and there is a lot more design conversions that go on.

The FPGA business for us is a newer business and one where we're kind of entering from scratch.

So, I do believe there will be different dynamics associated with that, regardless of the more rigor and the design end of definition.

Great, thanks a lot, Steve.

Hey, Mark, the gain on the buyback on the convertible for the quarter was $2 million.

Okay. $2 million.

So it sounds like then the balance was just your true run rate of interest income.

And when you look at your other income guidance for Q2, is that all based on true interest income?

That's correct.

Okay.

Thanks, guys.

We will now hear from Ruben Roy from Pacific Crest Securities.

Yes, thanks.

Jan, I'm wondering if anything changed in timing of your payments to Fujitsu?

No.

Okay.

So the milestones seem to be on track then.

And when do you plan on moving to production on the XP devices?

In the - - by the summer timeframe.

A final one for you, Steve, in terms of the 200 design ends that you were talking about.

Are there any specific end markets that are standing out that you are doing better in?

Or applications, if you could talk about that.

No, the product is really targeted toward the broad FPGA market.

Obviously we are focused on the low cost of segments.

So really the designs that we have secured reflect the business.

They are spread across geoographies, across applications, across the customer base.

And there's really nothing that I would call out as sort of a unique sector segment on any of those dimensions.

Okay.

That's all I have.

Thank you.

From EKN we'll hear from Larry Borgman.

Thanks.

In the communications sector that picked up, can you just highlight which segments of that were particularly strong or weak?

Sure.

Well, first, nothing was really weak.

We had a good quarter in communications.

The segment that stood out from a positive standpoint were the wireless segments, where we saw some good strength from designs in base stations.

And in addition, the access segments, those two were pockets of strength.

But generally, Larry, the communications segment was a bright spot for us last quarter.

Okay.

Just one other thing on the new products.

The nonvolatile FPGA where you are getting design wins.

Is there a particular type of customer that is focused on that as particularly useful to him?

Well, really, the XP family is a very differentiated FPGA.

It is offering really three things that we believe are important benefits to customers that are different than the SRAM FPGA's that are volatile that make up as I mentioned 90%-plus of the market.

The first of which is design security.

A programmable code that's implemented in a nonvolatile device is apparently much more secure than one that needs to be implemented into an SRAM device across the circuit board as the system powers up.

So, that's really the first benefit.

The second benefit is, with the nonvolatile technology, the programming code is actually inside the device.

And the chip can function upon power on.

And that's important to have logic functionality instantly in many applications.

Lastly, the ability to upgrade the chip in the field without interrupting the system operation is important for many applications.

So those are the technological benefits.

They tend to be relevant for areas where system reliability is important.

Where customers want to have kind of [5dyne] reliability to their system and have the ability to upgrade that in the field.

As well as areas where security is very important with respect to having proprietary logic implementations in the programmable devices.

And where OEM customers don't want competitors to have access to or potentially in the worse case, pirate or copy those algorithms.

So really those are the benefits we offer.

They are not unique to any application area, but we do believe are applicable to large portion of the market.

In fact, I would go so far as to say all things being equal, customers prefer a nonvolatile solution.

The trouble really is that historically, nonvolatile FPGA's have come at a substantial price premium versus the standard SRAM technology.

In fact, we would estimate that premium to be directly about three to five times that of an equivalent density SRAM device.

And despite that substantial premium, nonvolatile FPGA's currently make up about 8% of the market.

The LatticeXP is really revolutionary because combining an aggressive technology, I mentioned it is built on 130 nanometer co-developed imbedded flash process at Fujitsu.

If you couple that - - we have coupled that with an architecture that has been optimized for costs.

And it is basically the same fabric as the EC, our first platform FPGA is built on.

So with the combination of those two technologies, we can really price the LatticeXP, add parity to the price of the low-cost SRAM FPGA plus a BootProm.

So, really simple economics dictate that when the price comes down substantially, that should expand the market for nonvolatile FPGA's and that's what we intend to do.

Okay.

Thank you.

Bill Dezellem with Davidson Investment Advisors is next.

Thank you.

Steve, I would like to follow up just on your last comment.

That if the price premium for the nonvolatile FPGA has gone from three to five times that of an SRAM now to where there is price parity to the SRAM --

Bill, I didn't say price parity to the SRAM.

I said price parity to the SRAM plus a BootProm, a memory.

SRAM's again are volatile.

They need to have code put into them from some other nonvolatile source on the board.

One the techniques to do that is to use a boot memory to hold that code, and when the system powers up to send that code into the SRAM FPGA.

So you take the cost of those two discreet components, the FPGA and and the BootProm, that is what we are able to price at parity too.

Alright, and the nonvolatile FPGA - - let's see - - I am trying to figure out how much of a price reduction that equates to.

And therefore, what sort of a market potential that then leads to.

If in fact, at that three to five times price premium, there's still 8% market share.

It sounds like it's a very significant number over time.

It is a substantial price reduction versus the existing state-of-the-art nonvolatile technology.

Really, our viewpoint is that today the nonvolatile sector makes up about 8% of the market.

We think that can grow to 15% to 20% over kind of the four-year time horizon.

That's kind of how we think about this sector.

But, time will tell.

We're in the market now with a real product.

We are engaging customers and generating interest in this approach and this technology.

I think the most relevant, important thing for Lattice is that this provides some substantial differentiate.

It really brings some innovation and some benefit to the customer base.

It's a different approach and we hope to use that to build share in the FPGA market.

And then, Steve, a couple of additional questions.

Will R&D decline from its current level when the SC family is rolled out?

Or, what will be your expectation at that time.

Only if we stop developing the next generation products.

We need to continue to have a pipeline of new products that provide innovation to our customers to continue to grow our revenue.

So our expectation, Bill, is that R&D, as we have said many times should be relatively stable.

It will fluctuate depending upon kind of where we are with respect to new product releases.

Particularly with respect to mask and wafer expense for new product releases.

But from a head count standpoint and overall cost standpoint, we look for that to be reasonably stable.

And then last question is, to what degree are the new products - - the growth they are experiencing, is that cannibalizing your other products?

That - - cannibalization goes on all the time.

So really, a new product - - new products are made up broadly of new CPLD products and new FPGA products.

New CPLD products replace our older CPLD products.

And if we are doing our job, hopefully are cannibalizing our products and our competitors' products.

But most of that growth is replaced in business of our existing products.

And as the market grows and share changes, that can add to the overall revenue.

On the other hand, the FPGA business for us is very small, and today our new - - our FPGA products are very narrow in their market capture.

Our next generation FPGA's, which aren't producing any revenues today, they are producing designs and thus doesn't really contribute to all the new product revenue growth that I disclosed and we show in our press release every quarter.

The new FPGA products really expand the market and don't target the same areas that we are already in.

And for that matter, the revenue from our FPGA products, our first-generation products is relatively small.

We, we expect the LatticeEC the LatticeXP products to really be additive and the cannibalization minimal.

Great, thank you.

Our next question comes from Robert Toomey with RBC Dane Rauscher.

Good morning.

I have a question, Steve, THAT has to do with a statement you made regarding your guidance.

I think you said in your comments that you expect operating expenses to increase sequentially.

Whereas, in the press release I believe it says "decrease" by approximately 1 to 2 million.

I believe I said decreased.

And if I said increased, I apologize.

The press release is correct.

Our outlook is for operating expenses to decrease with a "D," $1 to $2 million on a sequential basis.

Okay, thanks for that clarification.

Also I just want to verify with respect to your second-quarter guidance, with respect to shipments and inventory, correction in the markets, that that guidance reflects end demand and not inventory snap-back?

I just wanted to clarify that.

Yes.

We believe that our revenue can grow in the low to mid single digit percentage points on a sequential quarterly basis due to positive market conditions based upon consumption and our new product growth.

Okay.

And can you comment on the current size and the growth of the FPGA market and the potential for the segment of the market that you are targeting with your new products?

Sure.

Last quarter I talked about the fact that the industry ended 2004 on a two-quarter revenue slide that would make it mathematically difficult to post a high level of market growth on a year-over-year basis, if we do the math.

That's kind of the bad news.

The good news is the dynamics that caused the two-quarter softness in our market, really a demand slowdown in Asia and a build-up of inventory; as I've commented appeared to be behind us.

I think it can be a relatively good year for programmable logic on kind of a sequential quarterly growth basis.

Last time I told you that we thought the FPGA market would be kind of flat to 5% up for the full year.

I believe if the industry can post normal sequential growth rates for the rest of 2005, we should be at the high end of that rate, perhaps a little bit over.

So, you know, we do expect the FPGA market to grow on a year-over-year basis.

But more importantly, we expect sequential growth in the next quarter.

And if trends continue to hold and we have a normal year throughout the year, the products that we have introduced really targeted towards the growing segments within the FPGA markets that are the low-cost, high-volume segment.

With key targets and the nonvolatile segment which I talked a great deal about [which will pick up].

So we do think that there should be positive consumption in the FPGA market.

And more important the segments where our devices are targeted should grow within the FPGA market.

And what was the high end - - the range on the high end, Steve?

5% plus.

Okay.

And then I wondered if you could comment on your agreement - - recent announcement of the Synplicity arrangement.

Is that any particular significance for you in the market or for the ability - - your sales of your new FPGA products or I guess what is the significance of that?

I should say.

Synplicity provides synthesis tools which are a key technology in order for customers to implement designs into FPGA's.

They basically break down the standard design code into a FPGA architecture.

So really the Synplicity agreement does two things.

It allows us to provide a very popular third-party synthesis package bundled with the Lattice software suite.

And so that's a value to the customers.

And allows a very useful technology to be coupled with our place, route-in map tools and user interface that allows access to our FPGA devices.

So it really broadens the acceptance and capture radius of our technology.

So, we are pleased to be able to continue to bundle the Synplicity tool with our proprietary software.

Second thing that the agreement does is a new thing, and that allows for enhanced support in Synplicity's high-end tool called SimplifyPro, which we were not fully supported in before.

This will really bring the full support, the full power the Synplicity tool that is available to many tens of thousands of logic designers to have visibility into and provide a very strong support for our new FPGA's's under that tool.

All things considered, it should ensure that we have a wide access to a broad base of FPGA users who are familiar with and standardized on the Synplicity product.

Great, thank you very much

Wells Fargo Tad LaFountain has our next question.

The 150 customers who have accounted for these 200 design wins.

Are these products giving you any help in penetrating new customers and opening up accounts that heretofore have been difficult to deal with?

They sure don't hurt.

And I think particularly the LatticeXP will help very much in that dimension tad.

Customers who are strongholds of our competitors really are open to seeking additional suppliers for FPGA.

Provided that that supplier provides something that is both different and necessary for the designs.

So to the extent that we provide that, the products are useful to open up new accounts and new relationships for Lattice.

And, you know, that's happening.

And I believe it will happen more with the LatticeXP architecture.

And does that involve extra marketing costs to overcome the stickiness of the customers' familiarity with the customers' design tools?

It involves effort.

Does it involve extra effort?

That's a relative question.

We have a broad and strong global sales channel.

Typically we have relationships with all these accounts on the PLD side.

So we are there anyway supporting the PLD side of the business.

That really involves targeting that existing sales channel and application for us in the field towards these FPGA designs within those similar customers.

Okay.

And final question.

Could you just give us an update on what's happening to the simple PLD product line?

It's small and irrelevant.

So not much is happening.

We are continuing to sell them.

And as our - - Altara and Xilinx both sell small pin count, small IO count PLD's, nobody ever asked them about that.

It is an irrelevant, small, declining business.

So less than 5% of sales?

We don't disclose revenue figures by product families.

We have many product families.

We don't ever talk about the revenue of each of those product families on a one-off-basis.

Okay.

Thanks.

We will now hear from Sumit Dhanda with B of A Securities.

Yes, good morning.

First question, operating expenses after the June quarter, do we see them growing a little bit slower than sales?

How should we think about that?

R&D flattish, SG&A should grow slower than sales barring kind of one-time events like we had in the first quarter as we have discussed.

The other question I had on your XP part, could you compare and contrast that versus the nonvolatile part that Actel has?

Sure.

Actel has made some aggressive announcements regarding low-cost flash FPGA's.

I think in general having two companies push a new technology particularly to differentiated disruptive technology can actually be a positive, not a negative.

Because it raises customer awareness more rapidly and serves to really validate that technology.

You have kind of two people instead of one person doing it.

Additionally, Actel tends to have their own account base which historically has been different from our account base and the account base of the dominant suppliers.

They kind of have their niche in the marketplace.

Our intention is - - I really think I said in response to the last question is the position the LatticeXP at the mainstream FPGA accounts, Which happen to be the accounts where we already have relationships via our PLD products and to position those against the SRAM FPGA's.

When we do compete with Actel we think we have a few compelling advantages.

First, our device is based on the industry standard for look-up table architecture.

Again it's similar - - the same fabric as the EC family.

And Actel's devices really aren't based on that architecture.

So, there's really tremendous leverage with customers who are already familiar with that architecture approach.

From a technical perspective, we believe we will have a significant programming time advantage when compared to the Actel approach.

And that's really critical with customers who want to erase and reprogram their logic in the field without a significant system disruption.

It is one of the advantages we offer versus SRAM, is that we can reconfigure the device without disrupting the system.

And if your programming time is very long, that advantage is very negated.

And lastly, based on - - at least what I have seen from the public product availability schedules, we believe we'll have a time to market advantage with our LatticeXP family.

So that's how I think we kind of stack up in a nutshell.

Just one follow-up.

From an architectural perspective, it is similar that your programmable element is a flash cell?

Or is it different?

We actually - - we have a chip that combines flash with basic SRAM logic.

So there is actually an imbedded flash block on the chip that is integrated that holds the code.

And the chip is loaded in a matter of a few microseconds upon boot-up.

So it is a totally different approach than the Actel approach.

Just one more follow-up.

Sorry about that.

But how do you shrink your dye size with that approach?

I guess is my question then.

And price your part in that parity with the mainstream SRAM FPGA solution?

Again to emphasize, I said our part will be priced parity with SRAM FPGA plus a BootProm.

So what we've done is basically - - we can get the same shrink dynamics as we migrate down the process technology curve for the SRAM fabric.

Integrating the BootProm, integrating the flash onto the block provides us some integration benefits.

So, that's how we shrink the device.

No real secret.

Okay.

Was there anything further sir?

No, thank you.

We will now move on to Richard Shannon with Piper Jaffray.

Hi, guys.

Just a couple of quick questions.

First of all, Steve, on your commentary regarding design wins.

You mentioned numbers of wins and customers.

Can you comment on the size of these design wins, especially relative to what you see in the past?

Do you see an opportunity to increase overall dollars per design win?

You know, not really at this point.

I shared with you some specific designing metrics.

However, translating those numbers into a specific time-based revenue forecast would mean making specific assumptions.

Particularly with the one you made, that you asked about.

But also about the percentage of those designs that move into volume production, the shape of the revenue curve.

And at this point, remember, we've really only really had two full quarters of selling under a belt.

It is a new business for us.

We just really have limited experience in that.

It is far too early for me to make any specific predictions or to develop a model based on actual data.

Because I don't have any at this point.

I am also really loathe to provide you a kind of bold sweeping revenue forecast based solely on inflated optimistic customer and field sales forecasts.

Which, I have.

As, sometimes in the past it has happened in our industry.

Perhaps after two or three quarters when we have some actual data, we can put together a data-driven forecast and answer these type of questions.

But I am really not in a position to do anything but speculate now.

We will follow up with that in a few quarters then.

Let's see here.

When you look at your EC and your ECDSP which you've had out there sampling with customers, for a couple of quarters, you mentioned.

I think it's generally held in the FPGA industry, that it takes roughly two years or so to go from a design win to a ramp in revenues.

Do you get the sense that that could happen any shorter period of time with your lower-cost EC and ECDSP products?

Not really.

I mean in general - - and I think we are on the public record, so I'll reiterate it, that our experience is the time from initial customer design into production revenue for any programmable logic project on average is about 18 months.

You can usually believe that and don't believe it will be any different for the EC technology.

Okay.

Fair enough.

And lastly, You answered a question earlier about some of the drivers within the communications market.

Is there any way that you can quantitatively or qualitatively discuss any of the components in terms of percentage of communications?

You mentioned wireless, some access and others.

We don't break that out.

And it is tough to break out because many customers are in multiple segments.

But essentially we think of a communication market as being a tripod of three kind of major sub areas: wireless, kind of the wireline portion and the third portion would be the data networking area.

Roughly those are all kind of equivalent as a percentage of our business.

And you know moves around from quarter to quarter.

But we don't quantify the data any finer than what we disclose.

Okay, ranking those as rough equals is useful.

So I appreciate that.

And that's all for me.

Thank you.

Thanks

[OPERATOR INSTRUCTIONS] Our next question comes from Danny Kuo with Bear Stearns.

Yes, could you address the bookings momentum you have seen in this quarter and how that compares versus the first month of last quarter?

Well, last quarter ended strongly for us and that's - - I think most people probably would have concluded that since we raised our guidance in March.

And actually came in a little bit higher than our range.

This quarter, we've had a nice start.

And we have a sequentially up outlook and the business thus far into the quarter supports that outlook.

And Jan, just a clarification on the OpEx decline.

This quarter is that mostly R&D driven or is very split evenly between SG&A and R&D?

Can you repeat that Danny?

Yes.

The decline we are expecting to see in OpEx this quarter, is that mostly R&D or is that mostly SG&A?

It is mostly actually in SG&A because we booked some expenses with securities litigation.

So is there a long-term target for you guys in terms of SG&A that we should forecast for next year?

It really depends on the revenue, Danny.

We would expect as revenue to grow, SG&A to grow slower than revenue and the percentage of revenue of SG&A to trend down.

Thanks.

[OPERATOR INSTRUCTIONS]

All right.

And gentlemen, we have no further questions.

So I would like to turn the call back over to you for any additional or closing remarks.

Thanks, everybody, for getting up early.

Anybody have any further questions, please give us a call at the usual number at the Company.

Thanks.

This concludes today's conference call.

You may now disconnect.